Target Registration Error minimization involving deformable organs using elastic body splines and Particle Swarm...

Area of Science:

• Medical Imaging
• Surgical Technology
• Computational Anatomy

Background:

• Precise target localization is a critical challenge in minimally invasive surgery.
• Existing methods may struggle with accuracy, impacting patient outcomes.
• Need for robust real-time solutions in surgical navigation.

Purpose of the Study:

• To evaluate the usability of elastic body splines (EBS) for minimizing target registration error (TRE).
• To present a Particle Swarm Optimization (PSO) approach for determining optimal EBS parameters.
• To demonstrate TRE minimization during specific respiratory phases in abdominal surgery.

Main Methods:

• Development of an elastic body spline (EBS) model.
• Application of Particle Swarm Optimization (PSO) for EBS parameter tuning.
• Experimental validation on 21 liver tumor patients undergoing abdominal surgery.
• Real-time implementation on a standard workstation.

Main Results:

• Successful minimization of target registration error (TRE) using EBS.
• Optimized EBS parameters achieved during phases of minimal free breathing motion (FRE).
• Method demonstrated effectiveness in liver surgery scenarios.
• Real-time performance validated on a standard workstation.

Conclusions:

• Elastic body splines (EBS) offer a viable solution for enhancing target localization accuracy in minimally invasive surgery.
• The PSO-based parameter optimization is effective for TRE minimization.
• The proposed methodology is suitable for real-time application in abdominal and liver surgeries.